CV142L

(2007). In, Davis, R.C., Luirigio, A.J., & Herman, S. (Eds.),

Victims of crime (3rd ed) (pgs. 9-34). Thousand Oaks,
CA: Sage Publications.

Developmental Victimology
The Comprehensive Study of Childhood Victimizations
DAVID FINKELHOR

n this chapter, I sketch the outlines of the

field of developmental victimology. It is a
field intended to help promote interest in and
understanding of the broad range of victimizations that children suffer from and to suggest
some specific lines of inquiry that such an
interest should take. In promoting this holistic
field, I contend that the problem of juvenile
victimization can be addressed in many of the
same comprehensive and conceptual ways that
the field of juvenile delinquency has addressed
the problem of juvenile offending.
The field of juvenile delinquency stands as
a monument to social science, one of its most

mature, theoretically and empirically developed domains. By contrast, despite substantial research on specific child victimization
topics such as child abuse or child sexual
assault, there is no similarly integrated and
theoretically articulated interest that characterizes the field of juvenile victimization. In
comparison to juvenile delinquency, juvenile
victimization has much less theory about
who gets victimized and why, much less solid
data about the scope and nature of the problem, many fewer longitudinal and developmental studies that look at the "careers" of
victimized children, and much less evaluation

Author's Note: The author wishes to thank Richard Ormrod and Stephanie Halter for their help with data
analysis and computation and Kelly Foster for her help with preparing this manuscript.
For the purposes of compliance with Section 507 of PL 104-208 (the "Stevens Amendment"), readers are
advised that 100% of the funds for this program are derived from federal sources (U.S. Department of
Justice). The total amount of federal funding involved is $353,233.

CRIME AND ITS IMPACT

to ascertain the effectiveness of policies and
programs that respond to juvenile victims.
These deficiencies are ironic for a variety
of reasons. For one thing, children are among
the most highly victimized segments of the
population (Hashima & Finkelhor, 1999).
They suffer from high ]rates of the same
crimes and violence adults do, and then they
suffer from much victimization specific to
childhood such as child abuse and neglect.
Second, victimization has enormous consequences for children, derailing normal and
healthy development trajectories. It can affect
personality formation, have major ~ n e ~ i t a l
health consequences, impact on academic
perfor~nance,and also is strongly implicated
in the development of delinquent and antisocial behavior (Margolin & Gordis, 2000). It
is clear that because of several factors, such
as children's special developmental vulnerability to victimization, its differential character during childhood, and the presence of
specialized institutions to deal with it (such as
child protection agencies), the victimization
of children and youth deserves both more
attention and specialized attention within the
larger fields of criminology, justice studies,
and even developmental psychology. This
chapter addresses a variety of issues: how to
define and categorize juvenile victimizations,
what is known about the epidemiology of
child victimization in broad terms, and how
victimization changes across the deveiopmental span of childhood.
DEFINITIONAL ISSUES
The interpersonal victimization of concern to
developmental victimology is a special kind of
negative life experience that stands apart from
other life events. This victimization can he
defined as harms that occur to individuals
because of other hunzan actors hehaving in
ways that violate social izorins. The human
agency and norm violation components give
victimizations a special potential for traumatic
impact. It is different from other stresses and

traumas, such as accideiits, illnesses, bereavements, and natural disasters. Even though
we sometimes refer to "victims of hurricanes,"
"cancer victims," or "accident victims," the
more general referent for the term uictinzizatioiz is interpersonal victimization. In interpersonal victimization, issues of malevolence,
betrayal, injustice, and morality are much
more present than is thc case for accidents, diseases, and natural disasters. To a large extent,
moreover, interpersoiial victimizations engage
a whole special set of institutions and social
responses that are missing in other stresses and
traumas: police, courts, agencies of social control, and other efforts to reestablish justice and
mete out punishments.
Although this area is the traditional domain
for the field of criminology, one reason why
traditional criminology may not have fully
explored its childhood dimensions is that child
victimizations do not map neatly onto conventional crime categories. Although children do
suffer from all the crimes that ad~iltsdo, many
violent and deviant behaviors by human actors
that harm children are ambiguous in their status as crimes. The physical abuse of children,
although techllically criminal, is not frequently
prosecuted and generally is handled by a different set of social control agencies from the
police and criminal courts. Peer assaults, unless
very serious or occurring among older children,
are generally ignored by the official criminal
justice system.
To encompass these complexities, I have
proposed that the victimization of children be
defined as including three categories: ( I ) conventional crimes in which children are victims (rape, robbery, assault), which I will call
"crimes"; ( 2 ) acts that violate child welfare
statutes, including some of the most serious
and dangerous acts committed against
children, such as abuse and neglecr, but also
some less frequently disciissed topics such as
the exploitation of child labor-which I will
call "child maltreatment"; and ( 3 ) acts that
would clearly he crimes if committed by
adults against adults, but by convention, are

Deuelopme~ztalVictimology
not generally of concern m the criminal justice system when they occur among or
against children. These would include sibling
violence and assaults between preadolescent
peers, and those that might be termed "noncriminal juvenile crime equivalents," which
I will call "noncrimes."
Each of these categories is a complex
domain, hut each has its stereotypical forms,
which sometimes help and at other times hinder thinking about the category. When the
public thinks of crimes against children, what
stands out are stranger abductions and
extrafamily child molestations, situations of
adults threatening children in which the
proper domain of protective and retributive
action is clearly the police, courts, and criminal justice system. When thc public thinks
of child maltreatment, they tend to think
of parents abusing or ~ieglecting parental
responsibilities, and the appropriate domains
of intervention are family courts, social

Figure 2.1

work, and mental health remedies. The public also is aware that there is noncriminal violence against children, and they think of peer
assaults, offenses that would be handled by
parents or school authorities.
Different as their stereotypes may be, however, these are not neat and distinct categories; there is substantial overlap. Child
maltreatment is sometimes treated as criminal, sometimes not (Figure 2.1). Child molesting, for example, is often considered as both a
crime and a child welfare violation. The same
act of peer assault that might result in an
arrest in one jurisdiction may be treated as a
"noncrime" for parents or school authorities
to sort out in another jurisdiction. Moreover,
there are normative shifts that are in progress
(illustrated by arrows in Figure 2.1). Sibling
sexual assaults once may have been viewed as
neither crimes nor child maltreatment, but
increasingly they are being handled by criminal justice and child welfare authorities. The

Child Victimization: Crimes, Noncrimes, and Child Maltreatment

CRIME AND ITS lMPACT

abduction of children by family members is
increasingly being viewed as both a crime and
child maltreatment.
The category of "noncriminal juvenile
crime equivalents" is one that often creates
confusion or draws objections. Some might
see it as a watering down of the concepts of
"victim" or "crime" to include acts such as
peer or sibling assault among children. But it
is difficult to deny the functional equivalence,
for example, between one adult hitting
another, say, in a bar, and one child hitting
another, say, on a playground. To st~tdyvictimization in a developmental fashion, we
must look at functionally equivalent acts
across the life span, even if the social labels
placed on the acts change as the participants
get older. The cultural assurnption is that
these acts are less serious or less criminal
when they occur at earlier ages. Whether and
how these acts are different should really,
however, be a matter of empirical investigation. When studied, violence between
youuger children has not been found to he
less physically or psychologically injurious
(Finkelhor, Turner, & Ormrod, in press).
Understanding the basis for the social construction of victimization across the span of
childhood should in fact be one of the kcy
challenges for developrnelital victimology.
An even more prohlematic type of juvenile
crime equivalent, however, is spanking and
corporal punishment, which is a form of violence (defined as acts intended to cause
physical pain) and would he considered an
assault among adults. But corporal punishment is tiat just typically viewed as minor victimization but is actually viewed as salutary
and educational by many segments of society.
Because our definition of victimization requires
the violation of social norms, forms of normatively accepted corporal punishment may
not qualify. However, there are signs that a
normative transformatiou is in progress
regarding corporal punishment (Greven,
1990). A majority of states have banned all

its forms in schools, several Scandinavian

countries have outlawed spanking even by
parents, and the American Academy of
Pediatrics has officially opposed spanking.
Social scientists have begun to study it as a
form of victimization with short- and longterm negative consequences (Strassberg,
Dodge, Pettit, & Bates, 1994; Straus, 1994).
Some have argued that it is the template on
which other violent behavior gets built.
Clearly, a developmental victimology needs
to take account of corporal punishment, and
spanking in particular, although it may
deserve individualized theoretical and empirical treatment.
Another somewhat problematic category
in developmental victimology concerns indirect victimizations, situations in which
children witness or are closely affected by the
crime victimization of a family member or
friend. These include children who are firsthand witnesses to spouse abuse (Jaffe, Wolfe,
&Wilson, 1990; Wolak & Finkelhor, 1998),
who are deprived of a parent or sihling as a
result of a homicide (Kilpatrick, 1990). or
who are present hut not injured in playground
massacres or the public killing of a teacher
(Nader, Pynoos, Fairhanks, & Frederick,
1990), all situations that have been studied by
researchers. Although indirect victimization
affects adults as well as children, the latter are
particularly vulnerable to effects, due to their
dependency on those being victimized.
Because most of the acts creating indirect victimizations are crimes, these situations could
he readily categorized in the "crime" category, but some, such as the witnessing of marital assault, also are treated as child welfare
violations in which the child is seen as a
direct, not indirect, victim.
A new domain in developn~entalvictimology in recent years focuses on the topic of
Internet victimization. Three kinds of diverse
offenses have been subsumed under this
ruhric: (1)Internet sex crimes and solicitations
for such crimes, (2) unwanted exposure to

Developmental Victimology
pornography, and (3) harassment and cyherbullying (Wolak, Mitchell, & Finkelhor,
2006). When adults solicit underage youth for
sexual activities or even online interactions, it
falls in the category of conventional crime. But
although youth receive an apparently large
quantity of online sexual solicitations, it is difficult to assess how much of this is from adults
and involves individuals who are aware of
the underage character of their targets. The
Internet has also created an enormous exposure of young people to inadvertent and
unwanted sexual material, but although offensive to many youth and parents, it is not yet
clearly defined as a crime or child welfare
problem, in part because the harm element has
not been clearly established. Harassment and
cyber-bullying appear to he fairly straightforward extensions of conventional bullying
behavior into the realm of electronic communication and are therefore the easiest to categorize. It is still early to fully understand how
the development of a large electronic communications environment will alter the conception of or risk for victimization.
Another problematic category for developmental victimology is the one that includes
mass victimizations, class victimizations, and
institutional and policy victimizations.
Warfare and generalized ethnic violence have
great impact on children. Because the main
agent of this impact is individual violent or
hurtful acts perpetrated by individual people,
this does not stray too far froill the class of
victimizations I am considering here. Children
victimized hy governme~ital or institutional
policies, however, are in a different domain.
Children deprived of rights or affected by
budget cuts or land expropriations or environmental policies are often seen as victims of
human agents, sometimes acting outside of
established norms. However, these are victimizations that fall far enough outside the
domain of the other interpersonal actions I am
callsidering within this field that they need to
be the subject of their own specialization.

An additional definitional complexity in

the domain of developmental victimology is
that, unlike in the domain of adult victimization, specific victimization categories have
been much less clearly drawn. Thus, for
example, child sexual assault, child sexual
abuse, and child molestation are often
thought of interchangeably, but these terms
also can refer to very different portions of
the problem of sexual offenses involving
children. Thus child sexual abuse, when discussed in child welfare contexts, often means
sexual offenses committed against children
by caretakers aud thus might not include
sexual assaults by strangers or peers. Child
molestation in colloquial terms is thought of
as sexual offenses committed against children
by adults and thus might excl~tdedate rapes
and sexual assaults by other juveniles. Child
sexual assault is sometimes taken in its literal
meaning to refer to violent and forceful sexual crimes against children and thus excludes
nonassaultive sexual crimes against children.
All this suggests that the field could benefit
from a great deal of definitional refinement
and organization.
DIFFERENTIAL CHARACTER
O F CHILD VICTIMIZATION
The discussion of how child victimization
should be defined highlights the fact that child
victimization differs from adult victimization.
Children, of course, suffer from all the victimizations that adults d e h o m i c i d e s , robberies,
sexual assault, and even economic crimes
such as extortion and fraud. But one salient
difference is that children also suffer from
offenses that are particular to their status. The
main status characteristic of childhood is its
condition of dependency, whicli is a function,
at least in part, of social and psychological
immaturity. The violation of this dependency
status results in forms of victimization, such
as physical neglect, that are not suffered by
most adults (with the exception of those,

CRIME AND ITS IMPACT

such as the elderly and sick, who also become
dependent).
The dependency of children creates what
might be thought of as a spectrum of vuinerability. Interestingly, the victimization types
that children suffer from can he arrayed on a
continuum, according to the degree to which
they involve violations of children's dependency status (Figure 2.2). At the one extreme
is physical neglect, which has practically no
meaning as victimization, except in the case
of a person who is dependent and needs to he
cared for by others. Thus it is a form of victimization that is created by children's dependent status and occurs primarily, if not
exclusively, to children. Similarly, family
abduction is a dependency-specific victimization, because it is the unlawful removal of a
child from the person who is supposed to be
caring for him or her. Other kinds of child
victimization are a hit more ambiguous.
Emotional abuse happens to both adults and
children, hut the sensitive psychological vulnerability of children in their dependent relationship to their caretakers is what makes
society consider emotional abuse of children
a form of victimization that warrants an
institutional response. Therefore it is fair to
say that emotional abuse is a dependencyrelated victimization as well.

At the other end of the continuum are

forms of victimization that are defined
largely without reference to dependency and
that exist in very similar forms for both
children and adults. Stranger abduction is
prototypical in this instance, because both
children and adults are taken against their
will and imprisoned for rausom or sexual
purposes. Homicide is similar: the dependency status of the victim does little to define
the victimization. In some cases, to be sure,
children's deaths result from extreme and
willful cases of neglect, hut there are parallel
instances of adult deaths resulting from
extreme and willful negligence.
One might think that most forms of child
victimization are either dependency related or
not. But ill reality, there are forms of child victimization that actually should he located
along the midsection of the dependency continuum. Sexual abuse falls here, for example,
because it encompasses at least rwo different
forms, one dependency related and one not.
Some sexual abuse entails activities ordinarily
acceptable between adults, such as consensual
sexual intercourse, that are deemed victimizing
in the case of children because of their immaturity and dependency. But other sexual abuse
involves violence and coercion that would be
victimizing even with a nondependent adult.

Negiect
Family
Abduction

Sexuai
Abuse

Physical
Abuse

Stranger
Abduction

Emotional
Abuse

Dependency-Related
Victimization Types

Figure 2.2

Dependence Continuum for Child Victimization Types

Not Dependency-Related
Victimization Types

Developmental Victimolofi)
In the case of physical abuse, there also is
some mixture of types. While most of the violent acts in the physical abuse category would
be considered victimizing even between
adults, some of them, such as tlie shaken
baby syndrome, develop almost exclusively
ill a caretaking relationship in which there is
an enormous differential in size and physical
control. The dependency continuum is a useful concept in thinking about some of tlie
unique features of children's victimizations.
It also is helpful in generating some hypotheses about the expected correlates of different
types of victimization at different ages.

SCOPE OF CHILD VICTIMIZATION

There is no single source for statistics on child
victimizations. The National Crime Victimization Sol-vey (NCVS), which is the ultimate
authority on crime victimization in general,
has two i~ilfort~inate
deficiencies when it
comes to child victimization. First, it does not
gather information on victims younger than
age 12. Second, it does not cover certain forms
of child victimization such as child abuse, sexual abuse, and kidnapping that preoccupy
p~thiicpolicy regarding children. But national
estimates that compensate for these deficiencies of the NCVS are available from some
other soui-ces. Some of these various estimates
are arrayed in Table 2.1.
Under some \~ictii~iization
categories, the
estimates of several different studies have
been listed, sometimes showing widely divergent rates. These differences stem fro111a variety of factors.
Some of the studies listed base their rates
on cases known to authorities (NCANDS) or
professionals (NIS-3). Such studies are cel-rain to count fewer cases than studies that
obtain information directly from youth and
their families. Wliile it misses many cases, the
advantage of i~lforination from authorities
and professionals, however, is that professional judgment is typically invol\~ed in

assessing whether a real qualifying victimization (e.g., physical abuse) occurred.

Other discrepancies are more complicated
to explain. For a variety of victimizations in
Table 2.1, estimates are available from
both the NCVS aiid the Developmental
Victimization Survey (DVS; Finkelhor et al.,
2005b), a sti~dyconducted by the author and
colleagues. The NCVS is a highly rigorous
survey conducted every year by tlie U.S.
Bureau of the Census, that interviews nearly
10,000 youth ages 12 to 17. The DVS was a
survey of both youth and caretakers regarding the experiences of 2,020 children from
the ages of 2 to 17. The NCVS estimates are
considerably lower than those from the DVS
for every crime and also lower than many
other survey estimates of specific forms of
juvenile victimization (Wells & Rankin,
1995). This is generally attributed to several
factors. The NCVS uses a complex definition
for each crime it measures, and respondents
need to endorse several sets of questions in
specific ways in oi-der to qualify. Second,
the NCVS interviews respondents on several
occasions over a period of three years to
iiiake sure that the incidents reported clearly
fail within and not outside tlie exact one-year
time period being investigated. Third, the
NCVS survey clearly orients respondents to
the topic of co~iventioiial "crinie," so incidents that respondents might not think of as
crimes (e.g., forced sex by a dating partner
or being beaten by a parent) may not get
reported. Fourth, the NCVS does not reqiiire
that youth be interviewed confidentiallj-, and
young people may fail to disclose incidents
they would not want their parents or family
members to know about.
What this means is that the NCVS estimates are very conservative and count primarily incidents that would be considered
conventional crimes in the narrow sense.
The DVS estimates, by contrast, are probably
inflated with minor incidents and incidents
that sor~ieobservers might dismiss as "not

Table 2.1

Rates and Incidence of V ~ t i o u sChildhood Victinlirations

Age
Assault, any physical

Witnessing!
Domestic violence
Family abductions (or
custodial interference)

grade
2-17

Soi6rceb

Repovt Type

DVS
hlCVS
DVS
DVS
NCVS
DVS
NCVS
2003
DVS
NCVS
DVS

SelflCaretaker report
Self-report
SelflCaretaker report
SelWCarrtaker report
Self-report
SelflCareraker report
Self-report

Agency reports
Agency reports
SelflCaretaker report
Self-report

2002
2002
1993
1995
2002
2002

NCANDS
NIS-3
DVS
Hostlie
llallways
DVS
NCANDS
NIS-3
CTSPC-Gallup
DVS
NCANDS

1993
1995
2002

NIS-3
CTSPC-Gallup
DVS

Agency reports
Self-reports
SelflCaretaker report

2002
2002

NCANIIS
DVS

Agency reports
SzlflCaretakei report

2002
1993
2002
2000

Sexual 11ar;issment

Psychological1
En~otionalabuse

Year

2002
1993-2003
2002

Sexual abuse
(rcxual assault by
known adult)

Neglect

No. Victimized

2002
1993-2003
2002
2002
1993-2003
2002
2004

S ~ b l ~ nassault
g
Rubbery

Phvsical abuse

Ratcl1000'

Notes

Nonsibling
Nonsibling

SelflCaretaker report
Self-report
SelfICaretaker report

SelfiCaretaker report
Agency reports
Agency reports
Self-reports
SelflCaretaker report
Agency reports

SelfICaretakcr report
Carctakrr reports

Includes
medical neglect

Age
Nonfamilv abductions

Homicide
Bullying

Source"

Report Type

0-1 7

NISMART-2

Caretaker reports

0-17

NISMART-2

Law enforcement

Online victimization
S e x i ~ a lsolicitations a n d
approachrs
Unwantcd exposure
t o sexual material
Harassment
Corporal punishnlent

Nu. Victimized

Year

0-1 7

SHR

6th-10th
grade

HBSC

Self-report

DVS
CTSPC-Gallup

SclfICaretaker report
Caretaker reports

2-17

DVS

SclfiCaretaker reDort

10-17

YISS-2

Self-reports

10-17

YISS-2

Self-reports

10-17

YISS-2

Self-reports

0-1 7

PCAA
ABC N c w s Poll

0-17

Notes
Legal definition,
includes
stereotypical
kidnappings
Stcrcotypical
kidnapping

Agency reports

0-17

2-17
Teasing o r
emotional bullying

Rate/1000"

Caretaker reports
Caretaker reports

Moderate and
frequent huilyilig

Spanked o r
hit ever

a. Numbers givcn in parentheses did riot appear in original source, but were derived from data piesflircd therein.

b. Source acronyms: DVS, Developmcntai Victimizarion Survey (Finhelhol-, Ormmd, Turner, & Hamhy, 200Sb); NCVS, National Crime Victimizarion Survey (Raum, 2005);
KCVS 2003, Nnrio~ralCiiriie Victimiration Survey, 2003 (Caralano, 2004); SCXNDS, National Child Abuse iY; Neglect Data System, 2002 (U.S. Department of Health and
llurnan Sei~,iccs-Adminisiracic~ri on C:biIdren Youth and l:amilies, 2004); NIS-3, 'niiid Nutiotial Incidence Study of Child Abuse and Neglect, 1993 (Sedlak & Rroadhurst,
19961; Hostile llallways (Axelrod & Markow, 2001); NISMART-2, Srcoiid Narional lricidence Study of Missing, Abducted, Runaway and Thiownaway Children, 1999
(ftammer, Finkelhor, & Sedlak, 2002; Sedlak, Finkelhor, liammer, & Scholtr, 2002); SHR, Supplemental Homicide Reports (For, 20051; HRSC, Health Behaviour of Schoolaged Children [Noilsei et al., 2001 ); CTSPC~Gallup(Straus, Hamby, Vinkelhor, Moore, & Runyan, 1998); YISS-2, Second Yoilth Internet Safety Survey (Wolak er al., 2006);
PCAA, Prcveiit Child Abuse America, (Onro, 1999); ABC News Poll iCiandall, 20021.

18
real crimes," such as sibling and peer assaults
and disciplinary acts. Table 2.1 reveals an
enormous quantity and variety of victimizations occurring to children and youth. Based
on the DVS, over half of all children experienced a physical assault in the course of the
previous year, much of it by siblings and
peers. One fifth experienced physical bullying,
and one fourth, emotional bullying. In addition, 1 in 7 experienced a theft, and 1 in 20 a
robbery. The NCVS rates are typically only a
fraction, in some cases a 10th or less of the
DVS estimates, which suggests how far we
may still be from a consensus about the epidemiology of child victimization. But even the
NCVS estimates suggest that conventional
crime victimization rates for youth are at least
three to four times larger than what is known
to police (Finkelhor & Ormrod, 2001) and
two to three times the victimization rate for
adults (Hashima Luc Finkelhor, 1999).

A TYPOLOGY OF CHILD
VICTIMIZATION BY INCIDENCE
The estimates for various types of child victimization, in spite of their methodological
limitations, definitional imprecision, and variability, nonetheless can be broken into three
rough and broad categories according to their
order of magnitude. First, there are the pandemic victimizations that appear to occur to a
majority or near majority of children at some
time in the course of growing up. Tl?ese
include, at a minimum, assault by siblings and
theft, and probably also peer assault, uandalism, and robbery. Second, there is what might
be called acute victimizations. These are less
frequent and occur to a minority, although
perhaps a sizeable minority, of children, hut
may he on average of a generally greater severity. Alnong these we would include physical
abuse, neglect, and family abduction. Finally,
there are the extraordinary victimizations that
occur to only a very sinall number of children
but that attract a great deal of attention. These

CRIME AND ITS IMPACT

include homicide, child abuse homicide, and
nonfamily abduction.
Several observations follow froin this
typology. First, there has been much more
public and professional attention paid to the
extraordinary and acute victimizations cornpared to the pandemic ones. For example,
sibling violence, the most fi-equent victimization, is conspic~lousfor how little it has been
studied in proportion to how often it occurs.
This neglect of pandemic victimizations
needs to be rectified. For one thing, it fails to
reflect the concerns of children themselves. In
a survey of children, three times as inany
were concerned about the likelihood of their
being beaten up by peers as were concerned
about being sexually abused (Finkelhor &
Dziuba-Leatherman, I995). The pandemic
victimizations deserve greater attention if
only for the alarming frequency with which
they occur and the illflirence they have on
children's everyday existelice. It is a rule of
public health that threats to well-being that
are minor or only have enduring consequences in a small number of cases can be
very serious in their total effects if they occur
frequently in a large population. So, peer
assaults could potentially, on
population
basis, be resl~onsiblefor more mental health
problems than child abuse.
Second, this typology can he useful in
developing theory and methodology concerning child victimization. For example, different types of victimization may r e q ~ ~ i r e
different conceptual frameworks. Because
they are nearly normative occurrences, the
impact of pandemic victimizations may be
very different from the extraordinary ones
that children experience in relative isolation.
Finally, the typology helps illustratc the
diversity and frequency of children's victimization. Although homicide and child ahuse
have been widely studied. tiley are notable
for how inadequately they con\.ey the variety
and true extent of the other victimizations
that children suffer. Almost all the figures in

Table 2.1 have been promoted in isolation at

one time or another. When we view them
together, we note that they are just part of a
torai environment of various victimization
dangers in which children live.
Poly-Victims
With so many victimizations occurring to
so many children, it is obvious that there must
be considerable overlap. Ironically, though,
the fragmentation of the field of child victimization has impeded inquiry into just how
niuch overlap there is and why. Advocates
and policymaliers concerned about one form
of child victimization or another, such as dating violence, have tended to present esti~iiates
and studies about their victims as though this
was the primary or only victimization that
such children suffered from. They could do
this hecanse studies of one kind of victimization rarely ask ahout other kinds. Some studies might inquire about multiple forms of
child maltreatment, such as physical and
sexual abuse. Other studies, like the NCVS,
inquire about multiple forms of conventional
crime, such as rape, robbery, and aggravated
assault. But studies almost never ask about a
very broad and coniprehensive range of victimizations, including child maltreatment,
conventional crime, and exposure to peer violence, for example.
It turns out that most jilvenile victims
experience muitiple victimizations. To ascertain this, we have developed a questionnaire
that asks ahout 34 different kinds of child victimization,
the
Juvenile
Victimization
Questionnaire. This questionnaire asks about
victimizations in five broad domains: conventional crime, child rnaltreatnient, peer and sibling, sexual victimization, and \vitnessing/
indirect victimization. This questionnaire was
utilized in a national survey of 2,020 American
children ages 2 to 17. So~neof the estimates
froln the survey, the Developlnent or
Victimization Survey, are listed in Table 2.1.

The survey found that victimization was a

frequent occurrence with 71% of the children
and youth experiencing at least one victimization in the last year. But more important,
it found the experience of multiple victimizations very common as well. We defined multiple victimizations as having a different kind
of victimization in a different episode over
the course of a year. This meant that an
assault and robbery on different occasions,
even by the same perpetrator, would count as
multiple victimizations, hut two assaults by
the same or even different perpetrators would
not count as multiple victimizations. This
conservative way of defining multiple victimization was adopted in light of findings that
different kinds of victimization seem to be
inore harmful than repeated episodes of the
same type (see Finkelhor et al., in press;
Fiitkellior, Ormrod, Turner, & Hamby,
2005a). Of the children with any victimization in the last year, two thirds had experienced two or more. The average number of
victimizations for a victimized child was
three in the last year, and the total ranged all
the way up to 15. Obviously, children who
had had one kind of victimization were at
increased likelihood to have other victirnizations as well. For example, if a child had been
physically assaulted by a caretaker, lie or she
was 60% more likely than other children to
also have been assaulted by a peer.
Children with multiple victimizations
should be of particular professional concern.
In other fields, it has been widely recognized
that multiple intersecting adversities frequently have impacts far beyond those of
individual stressful events. So, for example,
clients with several psychiatric diagnoses
(comorhidity) or who abuse different kinds
of drugs (poly-drug users] have been found to
pose particularly challenging problems.
There is every reason to believe that this is
also the case with child victims.
We have proposed to call this group of
multiply victimized children "poly-victims."

20
(We prefer to the term "poly-victim" over
"multiple victim" because the term "multiple
victim" can mean a victimization in which
there were several victims, a meaning that
could be confused with what we were intending to designate-a
victim who has had
several victimizations.) We expected that
research on poly-victims would show them to
be ~articularlyhighly victimized, vulnerable,
and distressed young people.
In fact, the DVS confirmed these predictions. We categorized as poly-victims the
youth in our national survey who had experienced four or more victimizations over the
course of the single year. Such youth comprised 31% of all victims and 22% of the full
sample. But they were the youth with the
most serious kinds of victimization. Forty
percent of the poly-victims had had a victiniizatiou injury, 42% had experienced a form
of maltreatment, and 25% had been victimized by a weapon-toting assailant. Although
they were not that different from other youth
in their demographic profile, they had considerably more other lifetime adversities,
such as major illnesses, accidents, or other
family problems. They were also clearly the
most distressed youth. They were 5.8 times
more likely than other youth to he angry,
20.2 times more likely to be depressed, and
10.3 times more likely to be anxious. In fact,
inost of the clinically distressed kids were
also poly-victims. For example, 86% of the
clinically depressed children also fit the criteria as poly-victims (Finkelhor et al., in press).
It appears increasingly that professionals
should be looking for poly-victimization
among children, not just one individual type
of victimization, even a serious one. Our analyses have suggested that poly-victimization is
most associated with mental health problems
and bad outcomes and that poly-victims are
the kids harboring the greatest amount of distress. The associatioi~sbetween distress and
individual victimizations disappear when
poly-victimization is taken into account

CRIME AND ITS IMPACT

(Finkelhor et al., in press). That is, children
who experience a single kind of victimization, such as bullying or even child maltreatment, appear to be able to recover from it.
But youth who experience victimization of
multiple kinds from multiple sources are
showing signs that they are locked in a pattern or trapped in a downward spiral that
should be of the greatest concern to those
trying to help.
As we come to understand poly-victims, it
may change some of the assumptions that we
have been used to making about victimization
in general. Victimizations have in the past
mostly been thought of as stressful or traumatic events. This is in part a legacy of the
field's close connection to the literature on
post-traumatic stress. The earliest victimization experiences to be studied in detail were
sexual assaults, which were considered to he
highly threatening individual episodes, happening to otherwise ordinary victims, who
were overwhelmed by a short-term incident.
But as victimization research has expanded,
we have come to understand that many victims are subjected to repeated episodes over a
period of time, as with the child who is hullied again and again on the playground or
emotionally and physically abused again and
again by a parent. We are also now seeing
that many children are subjected to a variety
of different kinds of victimization, such as
being beaten and sexually assaulted and
robbed, over a relatively short period of time.
This suggests that victimization for some
children is more like a condition than an
event. A condition is a mucli more stable and
ongoing process, whereas an event is more
time-limited. It is like the difference between
failing a test and failing a course, or the difference between an acute medical condition
such as appendicitis and a chronic one such
as diabetes. One of the most important diagnostic challenges that face professionals concerned about child victimization is discerning
those children for whom victimization has

become a condition, rather than just an

event. We should expect them to have different characteristics and a different prognosis.
Currently, most of what we know about
poly-victims is that they experience a lot
of victimization. They appear to be equally
divided between boys and girls, and they
appear to be somewhat more common
among older youth, although there are certainly considerable numbers of poly-victims
even at a very young age (Finkelhor et al., in
press). Current evidence does not strongly
suggest that they come from poor or minority backgrounds. Importantly, one feature
that does seein to he associated with polyvictimization is living in a family that has
been affected by divorce, separation, andlor
remarriage. Obviously, we need considerably
more study of these youth so that we can
identify them and prevent or remediate their
poly-victimization as a condition as early as
possible.

DEVELOPMENTAL PROPOSITIONS
Childhood is such an extremely heterogeneous
category4-year-olds and 17-year-olds having little in common-that it can he inherently
misleading to discuss child victimization in
general without reference to age. We would
expect the nature, quantity, and impact of
victimization to vary across childhood with
the different capabilities, activities, and environments that are characteristic of different
stages of development. This is the key principle of developmental victimology.
Unfortunately, the general culture is
already full of assumptions about development and victimization, many of them questionable and sometimes even contradictory.
Some victimizations are presumed to be
worse for younger children, others worse for
older children-mostly based on stereotype,
not evidence. We have already alluded to
some of these assumptions. Peer violence is
presumed to he more serious, injurious,

traumatizing, and crimelike as it occurs to

older children, for example. That is, a
teenager punched by another teen would be
regarded as experiencing something much
more serious than a five-year-old punched by
another preschooler. Is there evidence for
this? In fact, when we looked at these issues
in a research study, we did not find less injury
or psychological impact for younger children
in instances of peer violence (Finkelhor et al.,
in press). Still, they are not entirely equivalent
kinds of offenses if only because we have different mechanisms for responding to thempolice might want to arrest the teenage
assailant. But we probably should not assume
until we can study the matter more that the
acts are more dangerous or the consequences
more serious simply h e c a ~ ~ sthe
e participants
are older.
In contrast to peer violence, the colloquial
assu~nptionabout child molestation is that it
is inore serious for younger children. Some
people make the naive assumption that
because children are at an earlier developmental stage, they may be vulnerable to more
serious developmental disruption. For example,
a child who has not yet been introduced to
sex will be more affected by the molestation
than one who has developed some ideas and
concepts. But, here again, much of the available evidence casts doubt on the colloquial
assumptions. Some studies have found sexual
abuse and child molestatioii to have more
consequences at younger ages, and others
have found the opposite. One of the big prohlems is that victimizations that happen at an
earlier age tend to go on for a longer period
of time. It is clear that what developmental
victimology needs is a rigorously empirical
approach to developmental issues, one that
does not accept facile developmental assumptions at their face value. Things are generally
more complicated than most people, even
experts, presume.
One good place to start an empirical developmental victiniology is with propositions

22
about how the types of victimization and
types of perpetrators change over the course
of childhood. The mix of victimization types
is very likely to be different for younger
children and older children. Based on one of
the concepts introduced earlier, we v~ould
expect, for example, that victimizations stemming from the dependent status of children
should he most common among the most
dependent, hence, the youngest children. H
corollary is that, as children get older, their
victimization profile should come more and
more to resemble that of adults.
We can examine such propositions in a
crude way with the data that are available. In
fact, we d o know that some of the depeildency-related victimizations are most concentrated in the under-12 age group. For
examplc, physical neglect, the failure to take
care of the needs of a dependent child,
is heavily concentrated among younger
children. Family abduction is also heavily
concentrated among younger children. When
children are no longer so dependent, they
tend to make their own choices about which
parent to live with, and abduction is no
longer a feasible strategy for disgruntled
parents. By contrast, victimizations that we
grouped at the ~londependency end of the
contin~iuminvolve a greater percentage of
teenagers. Homicide is a crime defined equivalently for minors and adults, and it is concentrated ainong teellagers (Figure 2.2).
Homicide is a particularly good crime for
solne additional insights about devclop~nent
and victimization, because fairly coinplete age
data are available and because other efforts
have been made to interpret the patterns
(Christoffel, 1990; Christoffel, Anzinger, 8(
Amari, 1983; Crittenden 8( Craig, 1990;
Jason, 1983; Jason, Carpenter, iYi Tyler,
1983). Child homicide is also a complicated
crime fi-om a developmental point of view. It
has a coilspicuous bimodal frequency, with a
high rate for the very youngest children, those
uitder age 1, and another high rate for the

CRIME AND ITS IMPACT

oldest children ages 16 and 17 (Figure 2.3).
But the two peaks represent very different
phenomena. The homicides of young children
are primarily committed by parents, by choking, smothering, and battering. In contrast,
the homicides of older children are committed
mostly by peers and acquaintances, primarily
with firearms. Although the analysts do not
agree entirely on the number and exact age
span of the specific developmental categories
for child homicides, a number of propositions
are clear. There is a distinct group of neonaticides: children killed in the first day or few
weeks of life. The proportion of female and
rural perpetrators is unusually high in this
group (Jason et al., 1983). Homicide at this
age is generally considered to include many
isolated parents dealing with unwanted
children.
After the neonatal period, there follows a
period through about age five during which
homicides are still primarily committed by
caretakers using "personal weapons," the
criminologist's term for hands and feet, hut
the motives and circumsrances are thought to
he somewhat different from those pertaining
to the neonatal period. These preschool
victim homicides appear to he inostly cases
of fatal child abuse that occur as a result of
parents' attempts to control children or angry
reactions to some of young children's aversive behavior-uncontrollable crying, hitting
parents or siblings, soiling themselves, or
getting dirty (Christoffel, 1990; Ctitteilden &
Craig, 1990). Such children are freq~iently
thrown against hard surfaces, st]-uck hard
with a blow to the head or helly, or smothered. Because of their small size and physical
vulnerahility, many children at this age die
froin acts of violence and force by adults that
would not he fatal to an older child.
As children become school age, the rate
of child liomicide declines, and the nature of
child homicide becomes somewhat different.
Among school-age children, killings hy
parents and caretakers gradually decrease

Victim Age

I22 Family

E?B Acquaintance
.-

Figure 2.3

Stranger l
l
Unknown
~

.
...

Juvenile Homicide Rates, by Victim-Offender Relationship and Victim Age, 2003

SOURCE: SHK (Fox)2003.

and those by peers and acquaintalices rise.

There are more firearm deaths. Children get
murdered by suicidal parents bent on
destroying their whole families. Children this
age are also sometimes killed in the child
molesrations that begin to increase in this
period (although homicide is a rare accompaniment to child molesting). Some of the
children in this age group die as innocent victims in robberies and arsons. There is a mixture of the kinds of homicides that affect
younger children and also some of those that
affect older children, hut the overall rate is
low, and it is one of the safest times in the life
span in terms of homicide risk.
Then, at age 13, the homicide picture
changes again, and rapidly. The rate for
boys diverges sharply from that for girls.
Acquaintances become the predominant
killers. Gangs and drugs are heavily implicated for this group, and the rate for minority groups-African
Americans, Hispanic

Americans, and Asian Americans-soars.

The homicides for this group of youth look
a lot like the homicides for young adults,
although it is one of the few forms of victimization that they suffer at lower rates.
These patterns of homicide victimization
suggest some interesting propositions relevant to developmental victimology. First,
they suggest at least three somewhat different
"ecological niches" in which victimization
occurs: (1) a preschool, family-based, early
development niche (with a possible neonatal
subenvironment); (2) a middle childhood,
somewhat protected, mixed school and
family niche; and ( 3 ) an adolescent, riskexposed, transition-into-aduithood niche.
The types of homicide suffered by children
are related to the nature of their dependency
and to the level of their integration into the
adult world. Among the things that may well
change across childhood and across these
niches are the victim-offender relationship,

24

CRIME AND ITS IMPACT

the locale where the homicide occurs, the

nature of the weapons, the motives involved,
and the contribution victims make to the
crime in terms of risk taking and provocation. The homicide variations provide a good
case for the importance and utility of a developmeutal perspective on child victinlizations
and a model of how such an approach could
be applied to other types of victimization.
INTRAFAMILY VICTIMIZATION
Unlike many adults, children do not live alone;
they live mostly in families. hloreover, their
involvement in their families wanes as they get
older. So a plausible principle of developmental victimology is that younger children have a
greater proportion of their victimizations at
the hands of intimates and correspondingly
fewer at the hands of strangers. This is because
they live more sheltered lives and spend more
time in the home and around family.
Figure 2.4 indeed confirms this. Figure
2.4a shows data on crimes against children
known to the police from the FBI's National
Incident-Based Reporting System (NIBRS).

Family offenders are highest for the youngest

age victims. But the percentage declines
from near 70% to below 20% after age 32.
At the same time, acquaintance victimizations rise during childhood until adolescence,
where they plateau at about 70%. Stranger
victimizations remain low throughout childhood but start to increase a bit after age 15.
The patterns are very similar in data on victimizations reported in the DVS, shown in
Figure 2.4b.
These trends are verv consistent with what
we know about children's social development. Social activities expand throughout
childhood to include an increasingly large
and more distant network of contacts. But,
overall, children have fewer of the characteristics that might make them suitable targets
for strangers, such as money and valuable
possessions. In adolescence, they both
acquire such valuables and begin to interact
in even more public arenas so that increased
victimization at the hands of strangers makes
se1ise.
An additional possible principle is that the
identity of perpetrators may vary according

Police Data
80

3
\

Acquaintance

.....-.-.
._.-...--.-

Victim Age

-- .
- ...

Figure 2.4a

-. ..
-

Juvenile Victim Relationship to Offender by Victim Age: l'olice Dam

SOURCE: Fsdei.11 Bureau of Iniesrigarion (19971, National Iiicidrni-Rased Reporting Sysiem (NIBRS; 12 siatrs only),
computer file. Tahiilaiianr midenakcn by Crimes Againsr Cliildreii Research Center, Washington. D.C., U.S. Deparrnirnr
of Jusricr. Federal Bureau of Invrstigarion.

Survey Data

Victim Age

Figure 2.4b

Jovenile Relationship to Offendcr by Victim Age: Survey Data

SOURCE: De~~eiupmenral
Vicrimirarion Surrey. 411 forms of victimization (Finkelhor er al., 200Sh)

to the type of victimization and its place on

the dependency continuum. Victimizations
that are more dependency related should
involve more perpetrators who are parents
and family members. Available data suggests
that this is true. Parents comprise 100% of
the perpetrators of neglect (Sedlak, 1991)the most dependency-related victimizationhut only 28% of the perpetrators of homicide
(Federal Bureau of Investigation, 1992). This
pattern occurs because the responsibilities
created by children's dependency status fail
primarily on parents and fanlily members.
They are the main individuals in a position
to violate those responsibilities in a way that
would create victimization. Tlius, when a sick
child fails to get available medical attention,
it is the parents who are charged with neglecting the child, even if the neighbors also did
nothing.
Consiste~itwith developmental patterns in
victim-offender relationship and the dependency continuum, we would also expect that
more of the victiinizatio~ls of younger
children would take place in the hoine
and that victimizations would depart farther
and farther from the home as children age

and move out into an ever-widening circle of

social activity.
We would also expect that, as the homicide data shows, crimes against children
involving firearms would increase along with
development. In fact, one explanation for why
teens are murdered less than young adults in
spite of their equivalent or higher overall violent victimization rate could be that teens and
their associates have less access to firearms
than do young adults.

GENDER AND VICTIMIZATION

Developmental victi~nologyneeds to consider
gender as well as age in its effort to map the
patterns of victimization in childhood. In overall terms, many of the gender patterns seen
among adults also apply among children. Tliat
is, boys overall suffer more victimization than
girls, but girls suffer more sexual assaults. On
the basis of the conventional crime statistics
available from the NCVS and Uniform Crime
Report, the ratio of boys to girls for homicide
is 2.3 to 1; for assault, 1.7 to 1; and for robbery, 2 to 1. Girls suffer vastly more incidences
of rape (8.1 to 1; Bureau of Justice Statistics,

26
1992; Federal Bureau of Investigatioil, 1992).
But these ratios primarily pertain to the experience of adolescents, and they do not consider
age variations, which add a considerable wrinkle to the pattern.

THE AGE CRIME CURVE

The life course patterns in crime and delinquency have been one of the most interesting
threads for ongoing discussion and research in
criminology. The empirical foundation for the
discussio~l is the apparent observation that
criminal behavior accelerates dramatically
during the adolescent years to reach a peak in
young adulthood and then falls off in later
years. The dramatic rise from preadolescence
to adulthood has been ascribed to a variety of
factors. One argument is that it reflects a
biosocially based status competition for mates
that gets its start in adolescence (Kanazawa &
Still, 2000). Others contend that crime rises in
adolescence because at that stage young
people begin to have adult aspirations hut are
excluded from the labor market (Greenherg,
1985; Grogger, 2998). Others point simply to
the lax social controls that operate during
adolescence and young adultl~ood-singleli~~od,
no family responsibilities, and no cornmitmnlt
to employers. Does victimization risk have the
same age pattern, accelerating during adolescence in the same dramatic fashion as delin-I
quency? Official crime statistics would say yes,
but more comprehensive self-report surveys
suggest no.
Police data such as from NIBRS jurisdictions show that teens constitute three fourths
of the juvenile crime victims, with risk escalating as youth age (Finkelhor & Ormrod,
2000). Olily a few crimes, such as kidnapping, forcible sodomy, and incest, appear
more evenly distributed across developmental
stages. But the police data have serious limitations as valid testimony to the age curve for
victimization. Many of the victimizations of
younger children-assaults
at the hands of

CRIME AND ITS IMPACT

peers, abuse at the hands of parents, neglect
and other forms of child maltreatment-are
forms of victimization that are considerably
less likely to be defined as crimes or matters
of police concern.
The age patterns in victimization rates are
considerably different when the evidence
comes from victims themselves and their
family members, for example, from the DVS,
which assessed victimizatious from ages 2
through 17, using the same screening questions across all ages (Figure 2.5). Overall, victimization rose slightly but not precipitously
for the adolescents.' The rise was largest for
sexual victimizations and witnessingiindirect
victimizations. There was no rise for assaults.
Perhaps, most surprisingly, child maltreatment also rose with age. This might be the
form of victimization that we would most
expect to decline with age. In fact, some studies of child maltreatment known to professio~ials also show higher rates for older
children. But it may he the case that the maltreatment of younger children is difficult to
access, both in surveys (which almost of
necessity must get this information from the
caregivers themselves) and among cases
known to professionals, who are less likely to
have contact with younger children.
The absence of a steep increase in victimization is also apparent in the NCVS data.
Rates of violent crime measured in the NCVS
for 12- to 14-year-olds are as high as rates for
15- to 17-year-olds. Rape and aggravated
assault are a bit higher for the older adolescents, but simple assault is actually more
common for the younger youth. The steep
increases noted in self-reported delinquency
studies (Elliott, Huizinga, & Menard, 1989)
are not apparent in the self-reported victimization studies.
Why does the self-report information contrast so starkly with the official police data?
Studies clearly show that the younger the victims, the less likely it is that victimization will
be reported to law enforcement (Finkelhor 81

Developmental Victimology
Major Victimization Types by Victim Age

Wilnessllndirect
Assault
,-

,
,

Property

Maltreatment

---------

."

Sexual
____l__.___._^_.."_-----------------

10

11

12

13

14

15

16

17'

Victim Age (years)

Figure 2.5

Major Victimization Types by Victim Age

NOTE: Far puiposcs of geneiaii,ntioii, dar.3 tire shown

3s

Ormrod, 1999). The public and policc do not

want younger victims caught up in a judicial
system. They are less apt to define juvenile
victimizatio~isas crimes. Families, schools,
and child welfare officials lay claim to the
arbitration of offenses against younger victims. Younger victims themselves have a
harder time independently accessing police.
So, in spite of police data, victiiuization does
not accelerate in adolescence in the same way
as delinquency.

THE LIFESTYLES AND

ROUTINE ACTIVITIES THEORY
OF CRIME VICTIMIZATION
Developmental victimology also needs to
develop theories of victimization risk that take
into account the specific context of childhood.
This lnay mean altering some of the conventional approaches raken by victirnology as it
has been applied to adults. In victimology, in
general, one conceptual framework has dominated the discussioii: the closely related
"lifestyle exposure" and "routine activities"

-3-lrar iunniiig arcrages

theories (Cohen, 1981; Garofalo, Siegel, & I.auh,

i987; Cattfredson, 1986; Hindelang, Cattfredson,
& Garofalo, 1978). Such theories, as they
have been expoiliided in the past, highlight the
fact that lifestyles and activities of different
people put them in environments or situations
in which they are more or less in contact with
potential offenders and at risk of potential victimization.
Four central concepts have been used in
these approaches to explain the connection
between lifestyles and risk: proximity to
crime, exposure to crime, target attractiveness, and guardianship (bliethe & hleier,
1994). I'roximity to crime u,ould mean living
in high-crime areas. Exposure to crime would
i~icludethings s ~ ~ cash being out at night.
Target attracriveness would he attributes that
might entice offenders sucli as the ownership
of desirable and porrahle possessions. Gual-diaiiship highlights that spending considerable
time alone or apart from the family or other
possibly protective individ~~als
can create vulnerability. These concepts have proved useful
in explaining why certain groups such as

CRIME AND ITS IMPACT

men, blacks, and single people have higher
crime victiinization rates. They have also
been used to explain why rates of crime have
increased over time in some places and in
some periods, when, for example, fewer
people began living in families, and people
began acquiring moz-e conspicuously valuable
items.
When these concepts have been applied to
some extent to the analysis of youth victimization, it has been primarily to point out how
increased exposure and decreased guardianship heighten youth vulnerability. Young
people are viewed in this theory as engaging
in risky behaviors, such as staying out late,
going to parties, and drinking, which compromise the guardianship provided by parents
and adults and expose them to more possibilities for victimization (Jensen 8: Brownfield,
1986). Much of thc research on youth victimization has particularly stressed its connectioii
to delinquent activities (Lauritsen, Lauh, &
Sampson, 1992; Lauritsen, Sampson, 8: Laub,
1991). Delinquency is seen as a lifestyle that
puts a person in close proximity to other
offenders-aggressive or delinquent cornpanions or rival gang members. Moreover, it also
greatly reduces guardianship because delinquents tend to avoid conventional social environments and t h r o ~ ~ gtheir
h
activities also
largely forfeit their claims on the protection of
police and other authorities (Sparks, 1982).
Empirical research has confirmed that delinquents are indeed more prone to victimization
than otlier youth (Lauritsen et al., 1992;
Lauritsen et al., 1991).
CRlTIQUE O F LIFESTYLE AND
ROUTINE ACTIVITIES THEORY
The lifestyle theory perspective of youth victimization has ultimately been fairly narrow.
For one thing, many youth get victimized
without being involved in delinquency.
Delinquent activities are primarily the domain
of adolescents, particularly adolescent boys,

but even young children get assaulted, kidnapped, and sexually abused (Finkelhor &
Dziuba-Leatherman, 1994) without any connection to delinquent behavior. For another
thing, the lifestyle and routine activities
theories were designed for and have always
been best at explaining street cri~nesuch as
stranger assaults and robberies. But much of
youth victimization, especially of younger
children, occurs at the hands of acquaintances
and family members (Finkelhor & DziubaLeatherman, 1994).
These acquaintance and intrafamily victili1izations are not as well suited to the lifestyle
or routine activities concepts. For example,
routine activities studies often operationalize
exposure to crime as the amount of time routinely spent out at niglit or away from the family
household. However, when trying to explain
parental child abuse, such explanations collapse. It docs not increase a child's risk of
parental abuse to be away from their parents.
In fact, it may actually reduce it.
Thus it is not surprising that theories developed to explain children's victimization by
acquaintances and family members liave virtually ignored lifestyle theory and liave relied
on other concepts besides exposure and
guardianship. For example, in trying to account
for who becomes the target of bullying,
observers have noted that these tend to be
children with "avoidant-insecure" attachment
relationships with primary caregivers, who
lack trust, have low self-confidence, have physical impairments, are socially isolated, and
are physically weaker (Olweus, 1993; Smith,
Bowers, Binney, 8: Cowie, 1993).
The literature on physical abuse also
takes a very different tack from the lifestyles
approach. This literature tends to equate
victimization risk primarily with family and
parental attributes, such as family stress,
isolation, alcoholic and violence-prone caretakers, parents who have victimization
histories and unrealistic expectations of
their children (National Research Council,

~evelopnzelatalVictimology
1993), and youth characteristics such as
oppositiollal behavior, difficult temperament, or impairments that are a burdcn or
source of disappointment for caregivers
(Berdie, Berdie, Wexler, & Fisher, 1983;
Garbarino, 1989; Libby & Bybee, 1979;
Schellenbach & Guerney, 1987). A still different victimization literature, the one on
child sexual assault, notes even other risk
factors: girls, children from stepparent
families, children whose parents fight or are
distant and punitive, reduced parental
supervision, and emotional deprivation that
make children and youth vulnerable to the
offers of attention and affection that sexual
predatory offenders sometimes use to draw
children into sexual activities (Finkelhor,
1993; Finkelhor, 1994).
The concepts from these various literatures
can, to a limited extent, be subsumed into the
routine activities conceptual framework.
Tl~us,for example, lack of supervision (considered a risk for sexual abuse) corresponds to
the guardianship concept. Family social isolati011 (as a risk for parental physical abuse)
also has an element of missing guardianship,
but in this case the guardians are not the
family members themselves, but members of a
related social network. One might also coilsider characteristics such as having an impairment, being insecurely attached, being a
female, or being emotionally deprived as features of "target attractiveness."
But target attractiveness, in the routine
activities literature, has primarily been utilized in a very narrow sense, in reference to
the value and portability of material objects
that as a result of their lifestyle a person may
own or carry (Hough, 1987; Miethe & Meier,
1994). It could be extended without too much
distortion to refer to the value of a victim as
an object of desire, such as for a sexual crime.
But target attractiveness takes on a very different nleaning in the case of violent victimizations, one in which the word attractio,~
seems quite inappropriate. A child who is

beaten by a parent because the child's disability disappoints and frustrates a parent is an
"attractive target" for parental anger in only
a very ironic and convoluted way. Moreover,
it is not necessarily true, as is often the case
for property crime, that the offender is
simply choosing among more attractive targets, deciding to burglarize the home with the
fancier exterior. In the example of parental
assault, if the child were not disabled, it is not
clear that some other child would then suffer
the abuse illstead. Maybe, i this case,
nobody would be abused.
But perhaps thc biggest objection to trying
to subsume these child victimization risk factors into routine activities theory is that none
of these target attributes constitutes a
"lifestyle." Nor do they necessarily increase
risk through routine activities. Thus, femaleness, although it is a form of target attractiveness and does increase the risk for sexual
abuse, is not a routine activity. Moreover,
while malcness may put men at differential
risk for physical assault because men engage
in more unsupervised and risk-taking behavior (a lifestyle feature), femaleness does not
put women at differential risk for sexual
assault by virtue of anything they do. Femalcness itself is the risk attribute. Similarly,
while emotional deprivation may change a
person's routine activities, if a molester preys
on such a child because she is needy, it is not
the routine activities of the child that necessarily elevate the risk. The routine activities
idea of target attractiveness does not seem
broad enough.

A NEW CONCEPTUAL
FKAMEWORK FOR THINKING
ABOUT VICTIMIZATION
Thus, to explain the full range of victimizations among youth, the lifestyle or routine
activities framework needs to be modified.
Concepts such as guardianship, exposure, and
proximity, when it comes to victimization by

CRIME AND I'TS IMPACT

intimates, need to be seen not as aspects of
routine activities or lifestyles hut as environmental factors that expose or protect victims
from victimization. Thus, when a child is placed
at risk for sexual abuse because parents are
fighting and inattentive, the lack of guardianship is an environmental condition conducive
to victimization, m)t a problem of a lifestyle or
routine activity for the child.
In addition to the environmental conditions highlighted by the lifestyle theory to
explain the risks for youth victimization,
more attention also needs to be given to the
risk-increasing potellrial of individual characteristics and attrihutes, such as female gender
or emotional deprivation. These personal
characteristics of individuals would appear
to increase vulnerahility to victimizatiotl,
independent of any routine activities, because
these characteristics have some congruence
with the needs, motives, or reactivities of
offenders. 'That is, because certain offenders
are drawn or react to certain types of victims
or certain characteristics in victims, such victims are more vulnerable. This process might
be called "target congruence," and it increases
risk in one of three more specific ways,
referred to here as target t~trbzercrhility,target
gratzfiability, or target nrztagonism:

1. In the case of target vulnerability, some

victim characteristics increase risk because
they co~npro~llise
the potential victim's capacity to resist or deter victimizatio~land thus
make the victim an easier target for the
offender. For youth victimization. the pmtotypical risk factors in the vul~~erability
category would be attrihutes such as physical
u~eakness,emotional deprivation, or psychological problems.
2. In the case of target gratifiahiliq-, some
victim characteristics increase risk because they
are some quality, possession, skill, or attribute
that a11 offender wants to obtain, use, have
access to. or manipulate. The prototypical risk
factor in tlie gmtifiability category would he

female gender for the crime of sexual assault,

but keeping in mind that for some sexual
offenders, gratifiability focuses on prepubescent children or in some cases boys. Having
valuable possessions, as in the ro~iti~le
activities
notion of target attractiveness, would also fall
into this category.

3. In the case of target antagonism, some

characteristics increase risk by being qualities,
possessions, skills, or attributes that arouse the
anger, jealousy, or destructive impulses of the
offender. Examples in this category would be
ethnic characteristics or being gay or effeminate (for hate crimes), or being anxioi~sly
attached, a "mama's boy," etc. (as in the case
of b~illy victims). In the case of parental
assaults, characteristics such as being a burden
due to disability or being disobedient would he
other examples.
Although thcse target congruence concepts,
and particularly the target gratifiability one,
have similarities to the notion of target attractiveness, the word attrnct~venessand its stcreotypical applications in the crime of sexual
assault have victim-blaming connotations that
should be avoided. The attractions implied in
the co~lceptsused here are specific to the predispositions, proclivities, and reactivities of the
offender, hence the idea of congruence. Thus
gratifiability means tliat the target fits what the
offender is looking for, ~vhetherconventionally
desirable or merely satisfying of an offender's
idiosyncratic motive. Antagonism does not
i~uplyprovocation in the conventional sense:
without some predisposition, a crying baby
does not provoke assault any more than does
being the member of a minority.
It is important to note, as the ex.lmples
also illustrate, that target congruence changes
considerably from crime to crime, and from
offender to offender. Thus a female may have
more target gratifiability for a sexual assault,
but a male may have more target antagonism
for a gay-hashing. Characteristics that might
increase target antagonism for pal-ental

assaults, such as disobedience, may have little

if anything to do with risk for peer victimization. There may be some generalized target
congruence characteristics, such as weakness,
hut even this may be a relatively insignificant
factor in many victimizations.
These target congruence elements also
clearly play a greater role in some offenses
than others. In relatively impersonal street
crimes or group victimizations (e.g., sniper
attacks) and also in the case of family
members who live with very violent individuals, offenders may not he choosing victims on
the basis of any personal characteristic at all,
only proximity. In other victimizations (e.g.,
attempts to assassinate the president, stalking crimes, or a parent maltreating a colicky
habj-) the congruence of the personal characteristics of the victim \vitli the motives or
reactivities of the offender provide a virtually
complete explanation of victiiu choice.
t s to
These target congruence c o ~ ~ c e pseem
encompass most of the characteristics that
have been cited in the literature on youth
victimization outside the lifestyle theory
domain, characteristics such as low selfesteem and disobedience. But they also seem
qnite relevant to the prediction of forms of
victimization, such as street crime, which has
heen the primary focus of routine activities
research.
RESEARCH NEEDS
The reseiirch needs in the field of developmental viaimology are vast and urgent, given the
size of the problem and the seriousness of its
impact, and they range from studies of risk
factors to studies of treatment efficacy to stndies of criminal justice policy. But in the limited
space of this discussion, \ve will mention only
three important points.
First, if we are to take it seriously, we
need much better statistics to document and
analyze the scope, nature, and trends of
child victimization. The National Crime Victimization Survey records crime victimizations

only from age 12 and older. The Uniform

Crime Reports in the past have made no age
information available about crimes, with the
exception of homicide (something that is
changing under a new system, hiit the full
national implementation of this system is still
a long way off). The national data collection
system about child abuse also has scvere
methodological limitations, restricting the
way in which the information can be aggregated nationally or compared among states
(Finkelhor & Wells, 2003). We need comprehensive yearly national and state fignres on
all officially reported crimes and forms of
child abuse committed against children.
These need to he supplemeiited hy regular
national studies to assess the vast quantity of
unreported victimization, including family
violence 2nd child-to-child and indirect victimization. While there are methodological
challenges in such efforts, studies such as the
ones referenced in this chapter demonstrate
that this is feasible.
Second, we need thcory and research that
s
cuts across and integrates the v a r i o ~ ~forms
of child victimization. A good example is
the work on post-traumatic stress disorder
in children, which has been applied to the
effects of various victimizations: s e x ~ ~abuse,
al
stranger abcluction, and the witnessing of
lioinicide (Boney-McCoy- Kr Finkelhor, 199.5,
1996; Eth & Pynoos, '1 985; Terr, 1990).
Similar cross-cutting research could be done
on other subjects, such as what makes
children vulnerable to victimization or how
responses by family ~iiemhersbuffer or exacerbate the impact of victimization. To be truly
synthesizing, this research needs to study the
pandemic victimizations, not just the acute
and extraordinary victimizations, which have
been the mail1 focus in the past.
Finally, the field needs a more developmental perspective on child victimization. This wo~lld
begin with ail understailding of the ~ u i xof victimization threats that face children of different
ages. It would include the kind of factors that
place children at risk and the strategies for

C R I M E AND ITS I M P A C T
victimization avoidance that are appropriate a t
different stages of development. It also would
differentiate h o w children a t different stages
react t o and cope with the challenges posed by
victimization. It is ironic that until recently the
problem of children a s aggressors has h a d more
attention in social science than has children a s
victims, reflecting perhaps the priorities of the
adult world. It is encouraging that a s the needs
of children are more fully recognized, this halance is finally changing.

REFERENCES
Axelrod, A., & Markow, D. (2001). Hostile
halluays: Bullying, teasing, and sexual harassment in school. Washington, DC: American
Association of University Women Educational
Foundation.
Baum, K. (2005). Juvenile victimization a,zd of{
ending, 1993-200.3 (Bureau of Justice Statistics
Special lleport No. NCJ209468). Washingron,
DC: Office of Justice Pmgrams, U.S. Department
of Justice.
Berdie, J., Rerdie, M., Wexler, S., 8( Fishrr, B.
(1983). An empirical study of families involved
in adolescent inaltreatmcnt. San Francisco:
URSA lnstiture.
Boney-McCoy, S., 8( Finkelhor, D. (1995). I'rior
victimization: A risk factor for child sexual
abuse and for PTSD-related symptornatology
among sexually abused youth. Child Abuse si
Neglect, 19(12), 1401-1421.
Boney-McCoy, S., 8( Finkcihor, D. (1996). Is youth
victimization related to trauma symptoms a i d
depression after conrrolli~lgfor prior symptoms and family relationships? A longitudinal,
prospective study. Journal of Consrrlting and
Clinical I'sychology, h4(6), 1406-1416.
Bureau of Justice Statistics. (1992). National crime
survey. Washington, DC: U.S. Department of
Justice.
Catalano, S. M. (2004). Crivrinal victimizatiotz.
2003 (BJSLNCVS Report No. NCJ 205455).
Washingtoii, DC: U.S. Department of
Justice.
Christoffel, K. K. (1990). Violent death and injury
in U.S. children and adolescents. American
Journal of Diseases of Children, 144, 697-706.

Christoffel, K. K., Anzinger, N. K., & Amari, M.

(1983). Homicide in childhood: Distinguishable pattern of risk related to developmental
levels of victims. The American Journal of
Forensic Medicin. a,zd Pathology, 4(2),
129-137.
Cohen, L. E. (1981). Modeling crime trends: A
criminal opportunity perspective. ]ournal of
Research in Crime and Delinquency, 17,
140-1.59.
Crandall, J. (2002). Most say spankinx's OK by
parents but not by grade-school teachers. ABC
News and ICR international Communications
Research. Retrieved February, 10, 2006, from
http://abcnews.go.com/~ections/~~/DailyN
spankingp011021108.html.
Crittenden, P. A,,
Craig, S. E. (1990).
Developmental trends in the nature of child
homicide.Jowrna1 of interpersonal Violence, 5,
202-21 6.
Daro, D. (1999). Pirblic opinion and behaviors
regarding child abrrse preuenlio~z:1999 srrniey.
Chicago: Prevent Child Abuse America.
Elliott, D. S., Huizinga, D., & Menard, S. (1989).
Multiple problem youth: Deli~quency, substance use. and mental health problems. New
York: Springer-Verlag.
Eth, S., LYr Pynoos, R. S. (1985). Post-Trainmatic
S h e s Disorder in childyen: Progress in psychiatr),. Washington, DC: American l'sychiatric
Press.
Federal Bureau of investigation. (1992). Critne in
the United States, 1993: Uniform crime
reports. Washington, DC: U.S. Department of
Justice.
Finkelhor, 0. (1993). Epidemiological factors in the
clinical identification of child sexual abuse.
Child Abuse 6-Neglect, 17, 67-70.
t7inkeIhor, D. (1994). Current information on rhc
scope and nature of child sexual ahuse. The
Fzlt~~re
of Children, 4(2), 31-53.
Finkelhor, D., 8( Dziuba-Leatherman, J. (1994).
Victimization of children. American Psycholo@st,
49(3), 173-181.
Finkelhor, D., & Dziuba-Leatherman, J. (1995).
Victimization prevention programs: A national
survey of children's exposure and reactions.
Child Abuse & Neglect, 19(2), 125-1 35.
Finkelhor, D., & Ormrod, R. K. (1999). Reporting
crimes against jz~uetziles (Juvenile Justice
Bulletin No. NCJ178887). Washington, DC:

U.S. Department of Justice, Office of Juvenile

lustice: An annual review of research. 7,
Justice and Delinquency l'revention.
251-287.
Finkelhor, D., & Ormrod, R. (2000).
Greenberg, D. F. (1985). Age, crime, and social
Characteristics of crimes against jcrveniles
explanation. American Joctrnal of Sociology,
(Juvenile Justice Bulletin No. NCJ 1790.34).
91, 1-21.
Washington, DC: U.S. Department of Justice,
Greven, P. (1990).Spare the child: Tlge religious roots
Office of Juvenile Justice and Delinquency
of punishment and the ps)~chologicnlinzpact of
Prevention.
physical abuse. New York: Alfred A. Knopf.
Finkelhor, D., & Ormrod, R. K. (2001). Child
Grogger, J . (1998). Market wages and youth crime.
abuse reported to the police (Juvenile Justice
Jourrral of Labor Economics, 16(4), 756-791.
Bulletin No. NCJ187238). Washington, DC:
Hammer, H., Finkelhor, D., & Sedlak, A. J. (2002).
U.S. Department of Justice, Office of Juvenile
Children abdi~cted by family members:
Justice and Delinquency Prevention.
Natio~lalestimates and charactel.istics (Juvenile
Finkelhor, D., Ornirod, R. K., & Turner, H. A. (in
Justice Bulletin No. NCJ1964661. Washington,
press). Poly-victimization: A neglected compoDC: U.S. Departmmt of Justice, Office of
nent in child victimization trauma. Child
Juvenile Justice and Delinquency Prevention.
Abuse & Neglect.
Ilashima, I'., & Finkelhor, D. (1999). Violent
Finkelhor, D., Orinrod, R. K., Turner, H. >
&
I.,
victimization of youth versus adults in tlie
Elamby, S. L. (2005a). Measuring polyNational Crimz Victimization Survey. Journal
victiinization using the JVQ. C l d i Ahusr 6of lnterpersunal Violence, 14(8),799-820.
Neglect, 29(11), 1297-1312.
Hindclang, M. S., Gottiredson, M., & Gatof:~lo,J.
Finkelhor, D., Orrnnxl, R. K., Turner, 1-1. A,, &
(1978). Victims of pe?.sonal crime. Cambridge,
Hamby, S. L. (200%). The victimization of
MA: Ballinger.
children and youth: A comprehensive, national
Hough, M. (1987). Offenders' choice of targets:
survey. Child Maltreatment, lo(]), 5-25.
Findings from victim surveys. lorcrnal of
Finkelhor, D., Turner, H. A., & Omirod, R. K. (in
Quantitative Criminology, 3, 355-369.
press). Kid's stuff: The nature and impact of peer
Jaffe, P. G., Wolfe, D. A., &Wilson, S. K. (1990).
and sibling violence. Child Abuse & Neglect.
Cl7ildren of battered women. Newbury Park,
Finkelhor, D., & Wells, M. (2003). Improving
CA: Sage I'ublications.
national dara systems about juvenile victimiraJason, J. (1983).Child homicide spectrum. American
tion. Cljild Abtrse & Neglect, 27(1), 77-102.
/o~<mnal
of Diseases of Children, 137, 578- 581.
Fox, J. A. (2005). Uniform Crime Reports [United
Jason, J., Carpenter,M. M., &Tyler, C. W. (1983).
States]: Supplementary Homicide Repol-ts,
Underrecording of infant homicide in the
1976-2003 IConiputrr file1 ICPSK043.51-v1
United States. American Journal of Public
(pp. 11-22). Ann Arbor, MI: Intcr-University
Health, 71(2), 195-197.
Consortii~nifor l'olitical and Social Research
Jensen, G. F., 8: Brownfield, D. (1986). Gender,
[producer and distributor].
lifestyles. and victimization: Bcyond routine
Garharino, J. (1989). Troubled youth, trouhled
activity theory. Violence and Victims, 1, 85-99.
familics: The dynamics of adolescent maltreatKanazawa, S., & Still, M. C. (2000). Why men comment. In D. Cicchemi & V. Carlsoii (Eds.),
mit crimes (and why they desist). Sociological
Child maltreatmeizl: Theoq and research of
Theory, 18, 454-447.
the cairses and conseqzrences of child abuse and
Kilpatrick, D. G. (1990). Viole~zceas a precursor
neglect (pp. 685-706). New York: Cambridge
of women's substance abuse: The rest of the
University Press.
drugs-violence story. Paper presented at the
Garofalo, J., Siegel, L., & Laub, J. (1987). Schnol98th Annual Convention of tlie American
related victimizations among adolescents: rln
Psychological Association, Boston, MA.
analysis of National Crime Sutvey (NCS) narLauritsen, J. L., Laub, J. H., & Sampson, R. J.
ratives. Jorrrnal of Quantitative Criminology,
(1992). Conventional and delinquent activities:
3.321-338.
Implications for the prevention of violent
C;ottfredson, M. R. (1986). Substantive contribuvictiniizatinn among adolescents. IJiolence arzd
tions of victimization surveys. Crime and
Victims, 7(2), 91-108.

CRIME A S D ITS IMPACT

Lauritsen, J. L., Sainpson, R. J., & Laub, J. H. (1991).
The link between offending and victimization
among adolescents. Criminoloyy, 29, 265-292.
Libby, P., & Bybee, R. (1979). The physical ahuse
of adolescents. lournal of Social Issues, S.5,
101-126.
Margolin, G., & Gordis, E. B. (2000). The effects of
family and community violence on children.
Annual Review of Psychol(~gy,i l , 4 4 5 4 7 9 .
Miethe, T. D., & Meier, R. F. (1 994). C ~ i m and
r its
social coiztext: Totvard an i~ttegratedtheory
of offenders, viclirxs, and situations. Albany:
SUNY Press.
Nader, K., Pynoos, R., Fairbanks, L.., & Frederick,
C. (1990). Children's I'TSD reactions one year
after a sniper attack at their school. Anzerican
loirrnal of l'sychiatry, 147, 1526-1530.
Nansel, T. K., Overpeck, M., I'illa, K. S., lluan,
W. J., Simons-hlorton, B., & Scliridt, P. C.
(2001). Bullying behaviors among U.S. youth.
JAMA, 28.7(16),2094-2101.
National Research Council. (199.3). U,zderstaizding
child abuse atzd rteglcct. Washington, IIC:
National Academy Press.
Olweus, D. (1993). Bullies nn the playground: The
role of victimization. In C. H. Hart (Ed.),
Children of playg~or<nds:Research perspectives a n d applications (pp. 8.5-128). Albany:
SUNY Press.
Schellenbach, C. J., & Guernry, L. F. (1987).
ldentificatioli of adolrscmt abuse and future
intervention prospects. lournal ofAdolesce~?ce.
1 0(1), 1-12.
Srdlak, A. J. (1991). National incidence and preiialence of child ahuse ai1d iicglect: 1985Revised report. Kockvillr, MD: W~.stat.
Sedlak, A. 1.. & Broadhurst, D. D. (19961. 'Tl~ird
rratiorzai incidence sti~d),of child ai~irseand
neglect. Washington, DC: U.S. Department o f
Health and Human Servicrs.
Sedlak, A. J., Finlielhor, D., Ilamnier, H., &
Schultz, D. J. (20021. Niztio~~al
estin?ates o/'
missing children: An oiieri,ier*. (Juvenile Justice
Bulletin No. NCj1964661. Washington, DC:
U.S. Department of Justice. Office of ju\,enile
Justice and De1inqurnc)- Prrvmtion.
Smith, P. K., Bowers, L., Binney, V.. & Cowie, H .
(1993). Relationships of children involved in
bullylvictim problems at school. In S. Duck
(Ed.), Learning abosrt reliitionshif~s(pp. 184205). New-bury I'ark. CA: Sage Pubiicatioiis.

Sparks, K. F. ( 1 982). Crime and delinquency issues:

A monograph series. In R. G. Sparks (Ed.),
Research on victin~s of crime: Acconlpiishments, issues, and new directioizs. Rockville,
MD: National Institute of Mental Health, U.S.
Department of Health and Human Services.
Strassberg, Z., Dodge, K. A,, Pertit, G. S., &Bates, J.
E. (1994). Spanking in the home tind children's
subsequent aggression toward kindergarten peen.
Dec,elopmmt and P.sychopatholog, 6, 44.5461.
Straus, M. A. (1994). Bealing the dcvil oirt of them:
Corporal pu,tishment in Anzerican families.
New Yark: Lexington Books.
Stmus, M . A., Hamby, S. I.., Finkeihor, D., Moore,
D., & Runyan, D. K. (1998). Identification of
children maltreatment with the Parcnt-Child
Conflict Tactics Scales: 1)evelopment and psyl
chometric properties dam for a x ~ t i o n a sample
of American parents. Child Abirse i; Ne,~lect.
22(4), 249-270.
Terr, L. (1990). Too scared to cry. New York:
HarperlCollins.
l1.S. Departnient of Hrnlth and Human ServicasAdministr.~tion o n Children Youth a n d
Families. (20041. Child Maltrealn?e~rt2002:
Reports from the states to the Nalional Child
Abuse 6-Neglect Data Sgste?iz. Washington.
DC: U.S. Government Printing Office.
Wells, L. E., & Kankin, J. I<. (1995).Juvenile victimiration: Convrrgent validation o f altcrnarive measurelnents. Joarmal (if Rese'zrch in Crime,
.32(i), 287-307.
Wolak, J., & Finkclhor, D. (1998). Children
exposed to partner vioience. 131 1. I.. Jasinski &
1. M. Williams (Eds.), l'artircr ~'itiicnce: A
co,iip~eihzsivereview of 20 yerirs of research.
Thousand Oaks, CA: Sage I'ublicarions.
\Volnk, I., Mitcliell, K. J., & Finkclhor, D. (20061.
Onlir?e uictinziiatiorz: iyeriis later. National
Cmter for Missing and Exploited Children
Bulletin. #07-06-025. Alexanilria, VA.

NOTE
I. Figura 2..i shows the perccnrags of each age
cohort with any victimization or a n y specific type
of iii.timization, but it does not shoil- the total frcquency of victimizations. Hmoever, taking into
accoiinr \.icrimization frequeiicy-\vhich is roughly
the same a t all ages-clots nor change the shape of
the lines shown in Figurr 2.5.